Christopher M. Ingrassia

Theoretical Quality Assessment of Myocardial Elastography with In Vivo Validation

Myocardial elastography (ME), a radio frequency (RF)-based speckle tracking technique with one-dimensional (1-D) cross correlation and novel recorrelation methods in a 2-D search was proposed to estimate and fully image 2-1) transmural deformation field and to detect abnormal cardiac function. A theoretical framework was first developed in order to evaluate the performance of 2-D myocardial elastography based on a previously developed 3-D finite-element model of the canine left ventricle. A normal (control) and an ischemic (left-circumflex, LCx) model, which more completely represented myocardial deformation than a kinematic model, were considered. A 2-D convolu-tional image formation model was first used to generate RF signals for quality assessment of ME in the normal and ischemic cases. A 3-D image formation model was further developed to investigate the effect of the out-of-plane motion on the 2-D, in-plane motion estimation. Both orthogonal, in-plane displacement components (i.e., lateral and axial) between consecutive RF frames were iteratively estimated. All the estimated incremental 2-D displacements from end-diastole (ED) to end-systole (ES) were then accumulated to acquire the cumulative 2-D displacements, which were further used to calculate the cumulative 2-D systolic finite strains. Furthermore, the cumulative systolic radial and circumferential strains, which were angle-and frame-rate independent, were obtained from the 2-D finite-strain components and imaged in full view to detect the ischemic region. We also explored the theoretical understanding of the limitations of our technique for the accurate depiction of disease and validated it in vivo against tagged magnetic resonance imaging (tMRI) in the case of a normal human myocardium in a 2-D short-axis (SA) echocardiographic view. The theoretical framework succeeded in demonstrating that the 2-D myocardial elastography technique was a reliable tool for the complete estimation and depiction of the in-plane myocardial deformation field as well as for accurate identification of pathological mechanical function using established finite-element, left-ventricular canine models. In a preliminary study, the 2-D myocardial elastography was shown capable of imaging myocardial deformation comparable to equivalent tMRI estimates in a clinical setting.

Parameterization of left ventricular wall motion for detection of regional ischemia

While qualitative wall motion analysis has proven valuable in clinical cardiology practice, quantitative analyses remain too time-consuming for routine clinical use. Our long-term goal is therefore to develop automated methods for quantitative wall motion analysis. In this paper, we utilize a finite element model of the regionally ischemic canine left ventricle to demonstrate a new approach based on parameterization of the left ventricular endocardial surface in prolate spheroidal coordinates. The parameterization provided a substantial data reduction and enabled simple definition, calculation, and display of three-dimensional fractional shortening (3DFS), a quantitative measure of wall motion analogous to the fractional shortening measure used in 2D analysis. The endocardial surface area displaying akinesis or dyskinesis by 3DFS corresponded closely to simulated ischemic region size and 3DFS identified appropriate wall motion abnormalities during experimental coronary occlusion in a canine pilot study. 3DFS therefore appears to be a reasonable candidate for clinical tests to determine its utility in identifying and quantifying acute regional ischemia in patients. By linking state of the art finite element models to the clinically relevant framework of wall motion analysis, the methods presented here will facilitate formulation, in silico prescreening, and clinical testing of additional candidate measures of wall motion.

Dynamic Cardiac Information From Optical Flow Using Four Dimensional Ultrasound

Quantitative analysis of cardiac motion is of great clinical interest in assessing ventricular function. Real-time 3-D (RT3D) ultrasound transducers provide valuable three-dimensional information, from which quantitative measures of cardiac function can be extracted. Such analysis requires segmentation and visual tracking of the left ventricular endocardial border. We present results based on correlation of four-dimensional optical flow motion for temporal tracking of ventricular borders in three dimensional ultrasound data. A displacement field is computed from the optical flow output, and a framework for the computation of dynamic cardiac information is introduced. The method was applied to a clinical data set from a heart transplant patient and dynamic measurements agreed with physiological knowledge as well as experimental results

Tracking of LV endocardial surface on real-time three-dimensional ultrasound with optical flow

Matrix-phased array transducers for real-time three-dimensional ultrasound enable fast, non-invasive visualization of cardiac ventricles. Segmentation of 3D ultrasound is typically performed at end diastole and end systole with challenges for automation of the process and propagation of segmentation in time. In this context, given the position of the endocardial surface at certain instants in the cardiac cycle, automated tracking of the surface over the remaining time frames could reduce the workload of cardiologists and optimize analysis of volume ultrasound data. In this paper, we applied optical flow to track the endocardial surface between frames of reference, segmented via manual tracing or manual editing of the output from a deformable model. To evaluate optical-flow tracking of the endocardium, quantitative comparison of ventricular geometry and dynamic cardiac function are reported on two open-chest dog data sets and a clinical data set. Results showed excellent agreement between optical flow tracking and segmented surfaces at reference frames, suggesting that optical flow can provide dynamic “interpolation” of a segmented endocardial surface.

Angle-independent strain mapping in myocardial elastography 2D strain tensor characterization and principal component imaging

A current limitation of the implementation of myocardial elastography in a clinical setting is the difficulty of interpreting the one-dimensional strain maps due to varying strain values in the wall of the left ventricle (LV). In this paper, we demonstrate a robust angle-independent method for 2D myocardial elastography on simulated 2D ultrasonic images of a 3D finite-element analysis (FEA) model of the LV. Two FEA, a control and a regionally ischemic, canine left-ventricular models, were used and model states were obtained in increments and accumulated from end-diastole (ED) to end-systole (ES). Two- dimensional (2D) displacement in the myocardium was estimated between ED to ES. These estimates were good approximations of the FEA solution (rms errors of 0.18 mm for lateral displacement and 0.12 mm for axial displacement). The 2D symmetric strain tensor was calculated from the displacements and angle- independent principal strains were obtained using eigenvalue decomposition of the strain tensor. Principal strains in the myocardium have been shown to approximate normal strains with respect to an anatomical coordinate system (5). To test this angle-independence, displacements were obtained from two different orthogonally placed transducer locations. Principal strains were estimated from both locations and showed good correlation to the FEA solution. Rms errors between the FEA model and 2D elastography (2DE) estimation of principal strains from both transducer locations were 1.7% and 2.4% strain, respectively. Visualizing the transmural strain using principal strains greatly simplified their interpretation. Moreover, abnormal deformation of the ischemic region, which was difficult to observe with axial and lateral strains, was clearly visible in the principal strain images. In summary, the feasibility of 2D elastography estimation of myocardial displacement and strain was shown. In this paper, we propose the use of principal strains as a more useful tool in the visualization of abnormal wall motion and the detection of ischemia and other related heart diseases.

A Theoretical Performance Assessment Tool for Myocardial Elastography

The main purpose of this paper is to develop a theoretical tool in order to fundamentally characterize the performance of Myocardial elastography and identify the optimal parameters to be used for the more reliable detection of ischemia or infarction. A complete representation of the left-ventricular function throughout an entire cardiac cycle was previously demonstrated through the use of a 3D finite-element analysis (FEA) model. This FEA model together with an ultrasound image formation model is used here in order to test the performance of 2D myocardial elastography at distinct phases of the cardiac cycle and at different states of myocardium, i.e., normal and ischemic, based on in vivo canine data. A previously developed 3D finite-element analysis (FEA) model of a normal canine left ventricle with 80 nodes and 40 elements was used to simulate all main phases of the cardiac cycle. The axial and lateral displacements within multiple image (x-y) planes across the left-ventricular volume were iteratively calculated and corrected to reduce the decorrelation noise. Given the excellent agreement between the FEA solution and the elastographic strains measured in 2D over an entire simulated cardiac cycle, myocardial elastography proves to be a reliable technique for the accurate assessment of the myocardial deformation in 2D at distinct phases of the cardiac cycle as well as detection of the ischemic region. Preliminary in vivo results of a standard short-axis view in a canine myocardium are shown validating the performance assessment using the proposed model

Evaluation of Analytic Estimates of Ventricular Wall Stress Using the Finite Element Method

The importance of ventricular wall stress to cardiac function has been well-documented [1, 2], although accurate quantification remains a challenge. In this study, three popular analytic formulas for estimating circumferential wall stress were comprehensively evaluated to identify the conditions for which their use may be appropriate. In particular, the equations of Laplace [3], Mirsky [4], and Janz [5] are commonly used in the fields of cardiology and echocardiography; despite the inaccuracy of key theoretical assumptions, they have been attractive for their simplicity. For validation, we employed specialized finite element methods, developed specifically for cardiac mechanics applications [6], to compute regional wall stress in a series of model chambers having systematically varying geometric and material complexity. We limited our analysis to circumferential stress for consistency with the theoretical equations, and because of its relevance to cardiac mechanics.Copyright

Model-Based Screening of Wall Motion Measures for Detection of Ischemia in Three-Dimensional Cardiac Images

Quantitative measurement of left ventricular wall motion can improve clinical diagnosis by providing a more objective approach than qualitative analysis, which is subject to large inter-observer variability. We have developed novel techniques for quantifying left ventricular wall motion in three-dimensional image data sets. In this study, finite element models simulating regional ischemia in the left ventricle were used to screen potential wall motion measures for their capability to detect and evaluate the size of an ischemic region. Preliminary experimental results showed that wall motion analysis of real-time three-dimensional echocardiographic images successfully detected ischemia. Our four-dimensional wall motion analysis system provides an objective and quantitative approach for detecting and assessing the severity of disease